Author Topic: Analog video capture device comparison  (Read 14711 times)

0 Members and 1 Guest are viewing this topic.

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Analog video capture device comparison
« on: January 06, 2021, 02:27:54 am »
Most thermal imagers have an analog video output which is often the only signal which can be easily interfaced.
In this thread I want to gather different solutions how this signal can be captured.

Each device will get it's own post with data, test images and video files.
I'll test with NTSC and PAL video from different sources and update the list with links to the devices.
If you want your own device in this list just post all the information you have here in this thread  :D

PC based capture

And some portable/standalone devices

Some information is already available in different threads:
Hauppauge WinTV, EzCap, ARCHOS mini PVR, Grandstream GXV3500
EZCap
DFG/USB2pro
ImmersionRC Power Play FPV DVR

Stay tuned  :D
« Last Edit: January 10, 2021, 06:23:14 pm by Cat »
On the Internet, nobody knows you're a cat.
 
The following users thanked this post: Fraser, _Wim_, basecase, svgurus

Offline Vipitis

  • Frequent Contributor
  • **
  • Posts: 867
  • Country: de
  • aspiring thermal photography enthusiast
Re: Analog video capture device comparison
« Reply #1 on: January 06, 2021, 09:42:51 am »
will be interested in the results. I might change my setup. I am really really unhappy with the image quality of the PowerPlay FPV. I contacted the manufacturer and they wouldn't tell me much about the encoder chip used or give me access to compile my own config file. The bitrates are really low and the interlacing is also not working great. I can supply example data from both MIRICLE cores with new and old firmware. however there are a lot of oddities attached which I really dislike. Will have to take some time to make a proper submission here later on. Is there anything specific you are looking for or got a format template?

I might also do a teardown to find what electronics are inside and might change the form factor for my device...
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 13170
  • Country: gb
Re: Analog video capture device comparison
« Reply #2 on: January 06, 2021, 11:53:42 am »
Cat,

This is a really welcome Post.

I have ‘played’ with several video capture devices over the years and often found them to be frustrating due to image quality issues, driver instability on particular operating systems or just plain junk that does not work !

My first ever capture solution was the ATI “All In Wonder Pro” combined PC video and capture ISA card. It did what it was supposed to do and, for its time, produced excellent captures. I then moved onto a Pinnacle PCI card and that worked well with no issues. It was only when I moved to the convenience of USB connectivity that the trouble started !

Some general comments on USB capture devices from my previous experiences.....

1. There are USB 1.1 and USB 2.0 versions commonly available. USB 2.0 should be considered a minimum speed and newer USB 3.0 are likely better for higher resolution, high frame rate captures. It just comes down to the maths of moving all those pixels into the host PC with the overhead of USB packet transfer to be considered. Poor transfer rates can lead to dropped frames and sometimes a capture card will reduce capture quality if the USB data rate is poor. Hence why USB 2.0 or better devices are a good idea. FireWire was an excellent data link for video capture but is now obsolete. USB3.0 is fast enough for demanding tasks.

2. It is important to discover the chipset that is used in a USB capture device. Whilst it may be new and shiny on the outside, it can contain an old and relatively poor performance chipset. Drivers can become an issue for some older chipsets, especially for Win10. The newer chipsets offer ‘system on a chip’ highly integrated video capture and have made the complex process of capturing video into a PC simple for hardware manufacturers. The chip takes video of many types in on one side and spits out USB packets on the other side  ;D This thread may help us to identify which chipsets perform well, and which are to be avoided.

3. The small USB video capture dongles often have a sealed case and no ventilation. They can get quite hot whilst in use and this does not bode well for longevity. If you do a lot of video capture work, there may be justification in fitting a better case to the dongle that provides ventilation. Alternatively, some ventilation holes may be added to the original case. Some USB TV capture dongles had ventilation slots and that seems a sensible idea to me. Chips that are working hard and getting too hot can misbehave, throttle their speed or even hang.

4. There are some VERY cheap video capture dongles available on eBay etc. In my experience these use the cheapest available chipset and components so do not expect such a unit to provide the best capture performance. Hence my warning about old chipsets and USB 1.1 data links. It is not only the video capture image quality that can be poor, it can also be the frame rate, available capture resolution and the occurrence of many dropped frames as the chipset struggles in its task to convert the analogue TV signal accurately into the digital domain and packetise it for transfer to the host computer. This is not to say you must pay a lot of money to get a decent USB video capture dongle. There are many reasonably priced solutions that will perform well without breaking the bank. There have been excellent advances in video capture silicon in recent years, hence my warning to maybe avoid older chipsets unless they are known to perform well. In my experience, the maximum capture resolution and frame rate can be a hint at the performance of the chipset. 320 x 240 at 15fps = one to avoid ! A more capable chipset will offer at least 640 - 480 pixels with a choice of frame rates to suit the requirements of the user. My old Matrox MXO2 Mini from 2010 offers up to 1080p at a range of high frame rates including the 24fps used for film transfers at the correct speed. That unit was very expensive in its day but there must be similar available now and likely at a much reduced cost. I cannot recommend the Matrox MXO2 as it is long out of support and was effectively abandoned at Windows7.

5. Connectivity ...... a video capture card or dongle can offer a range of video inputs (and outputs in some cases). The simplest capture devices use a composite video signal. This is the lowest quality signal that you will find on AV equipment. That said, it is likely what you will have to work with on thermal cameras that provide an analogue video output ! For information, in terms of quality this is the list from worst to best.....

Composite Video - Y/C (S-Video) - RGB - Component - HDMI

SDI will be in the area of HDMI I suspect and is found on professional camera equipment.

6. Monochrome video - This week I used an EZCap USB 2.0 dongle to capture video from a monochrome output SWIR camera. The dongle is actually a decent model but it did not like the monochrome video input and claimed no signal was present when I tried to record. I fooled the units input checks using a colour bar chart video input from my CCTV test monitor before swapping the signal to the camera output. It was a faf to do and highlights that modern video capture chipsets MAY have an issue with missing parts of a signal that they expect to find, such as the colour burst. Video capture devices offer several Television Signal standards in their input configuration but on my dongle all were colour and there was no option for a monochrome PAL or NTSC signal. I know that other capture dongles do not have this issue so it may be uncommon or limited to a certain chipset. Some may not check the video input for anything except the sync pulses.

7. Video standards conversion...... thermal imaging cameras commonly present either a PAL or NTSC signal at their video output socket. The video capture device needs to be set for the video standard that is present. All of my capture devices support PAL, NTSC and SECAM without issues. The video standard can be changed in software when the recording is transferred to a DVD, Memory Stick etc. There is not normally a need to change the video standard with a hardware solution between the camera and the video capture device.

8. Video link standard conversion..... this is a new one for me so I can offer no in depth experience beyond a single test I carried out recently. I had a DVD recorder that provided only component, SCART RGB and SCART Composite video output. I needed an HDMI signal to feed into our new television. I bought a “Techole” SCART to HDMI converter from Amazon. It had good reviews. I installed the converter and was impressed that it was truly ‘plug and play’ with only the output resolution of 720P or 1080P to be selected with a push button. The unit up scaled the input video signal :) Now the bad news. Whilst the converter worked well, I did some research and discovered that most inexpensive SCART to HDMI converters use a chip set that only accepts a composite video input, even if they have a SCART socket. This means that a video source that can provide RGB with better quality cannot deliver that improved image to the televisions HDMI input. More expensive converters will take component or RGB video inputs. Again, with a thermal camera the issue is moot as the output signal is often plain old composite video. In my case, I ‘dumped’ the elderly DVD recorder and bought a wonderful used Sony RDR-HXD870 HDD/DVD recorder that provides an upscaled 1080P HDMI output  ;D Great recordings and great image quality on our new TV.

Using a camcorder as a capture device ...... For many years the UK camcorder market offered Camcorders that could only output line video and not record it. I believe this was due to the different Duty rates between line out only camcorders and video recorders as a camcorder with a video line input was classed as a VCR. You could buy the same model of camcorder overseas that had the video line input capability enabled ! I bought such camcorders but, in truth, rarely used the line input recording feature. My last camcorder was a Canon DV tape based unit and that had the line input recording enabled. The Camcorder could record analogue video from a video source and play that digitally recorded video back via its FireWire port to a host PC for editing etc. It could also do live real time conversion from analogue video source to DV output from its FireWire port. Is this a good option in place of a modern USB video capture device ? Well it is a solution but there are issues. The quality of video conversion may, or may not be as good as that of a dedicated USB video capture device. The resolution of video capture may also be relatively low. The video standard of the input to the camcorder may be fixed so not suitable for other video standard recordings. The Camcorder may have an inactivity shutdown feature that switches it off if not actually recording. FireWire is sadly now obsolete so less common on modern PC’s and laptops. It can still be added via a PCIe card or Expresscard however. The camcorder is a more bulky and complicated solution than buying a decent dedicated video capture device.

Well I think that is enough from me for now. I will correct the likely many typos in the above text later  ;D

Fraser



« Last Edit: January 06, 2021, 03:29:15 pm by Fraser »
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 
The following users thanked this post: svgurus

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Re: Analog video capture device comparison
« Reply #3 on: January 09, 2021, 04:10:52 am »
Sorry for the late reply, had some minor technical difficulties (laptop HDD died :-BROKE)
Anyway, the first results are in with a FLIR Scout III 640 (Tau2) as signal source. During boot up the small ® Symbol at the FLIR logo is a nice feature to test the resolution of the capture device. Afterwards the FLIR logo in the top right corner and other symbols can be used to check the capture devices for cropping.

Unfortunately it seems like the analogue video output is NTSC only (found no setting in the FLIR GUI, maybe only the Scout III 320 can be set to PAL?).
Because of this all the current tests are done with NTSC. Why do I want to test the devices with PAL, too? Sometimes there are cropping and interlacing issues which can be different between the two video standards.

I have ‘played’ with several video capture devices over the years and often found them to be frustrating due to image quality issues, driver instability on particular operating systems or just plain junk that does not work !
My first ever capture solution was the ATI “All In Wonder Pro” combined PC video and capture ISA card. It did what it was supposed to do and, for its time, produced excellent captures. I then moved onto a Pinnacle PCI card and that worked well with no issues. It was only when I moved to the convenience of USB connectivity that the trouble started !
I had somewhat similar experiences, but even with the PCI cards there were huge differences in the image quality between different chipsets.

Some general comments on USB capture devices from my previous experiences.....

1. There are USB 1.1 and USB 2.0 versions commonly available. USB 2.0 should be considered a minimum speed and newer USB 3.0 are likely better for higher resolution, high frame rate captures. It just comes down to the maths of moving all those pixels into the host PC with the overhead of USB packet transfer to be considered. Poor transfer rates can lead to dropped frames and sometimes a capture card will reduce capture quality if the USB data rate is poor. Hence why USB 2.0 or better devices are a good idea. FireWire was an excellent data link for video capture but is now obsolete. USB3.0 is fast enough for demanding tasks.
Caveat emptor: Not all (or even most) capture devices marketed as USB3 are real USB3 devices, for example the HDMI grabber I used is also available as USB3 version but with the same chipset which does not have the additional data lines for USB3.
2. It is important to discover the chipset that is used in a USB capture device. Whilst it may be new and shiny on the outside, it can contain an old and relatively poor performance chipset. Drivers can become an issue for some older chipsets, especially for Win10. The newer chipsets offer ‘system on a chip’ highly integrated video capture and have made the complex process of capturing video into a PC simple for hardware manufacturers. The chip takes video of many types in on one side and spits out USB packets on the other side  ;D This thread may help us to identify which chipsets perform well, and which are to be avoided.
Exactly, this is why I wanted to create this thread and supply as much information as possible  :D

3. The small USB video capture dongles often have a sealed case and no ventilation. They can get quite hot whilst in use and this does not bode well for longevity. If you do a lot of video capture work, there may be justification in fitting a better case to the dongle that provides ventilation. Alternatively, some ventilation holes may be added to the original case. Some USB TV capture dongles had ventilation slots and that seems a sensible idea to me. Chips that are working hard and getting too hot can misbehave, throttle their speed or even hang.
Exactly. But as thermal imaging enthusiasts we can easily improve the devices with additional cooling if necessary.

4. There are some VERY cheap video capture dongles available on eBay etc. In my experience these use the cheapest available chipset and components so do not expect such a unit to provide the best capture performance. Hence my warning about old chipsets and USB 1.1 data links. It is not only the video capture image quality that can be poor, it can also be the frame rate, available capture resolution and the occurrence of many dropped frames as the chipset struggles in its task to convert the analogue TV signal accurately into the digital domain and packetise it for transfer to the host computer. This is not to say you must pay a lot of money to get a decent USB video capture dongle. There are many reasonably priced solutions that will perform well without breaking the bank. There have been excellent advances in video capture silicon in recent years, hence my warning to maybe avoid older chipsets unless they are known to perform well. In my experience, the maximum capture resolution and frame rate can be a hint at the performance of the chipset. 320 x 240 at 15fps = one to avoid ! A more capable chipset will offer at least 640 - 480 pixels with a choice of frame rates to suit the requirements of the user. My old Matrox MXO2 Mini from 2010 offers up to 1080p at a range of high frame rates including the 24fps used for film transfers at the correct speed. That unit was very expensive in its day but there must be similar available now and likely at a much reduced cost. I cannot recommend the Matrox MXO2 as it is long out of support and was effectively abandoned at Windows7.
Some of the devices I tested were the cheapest ones available :D

5. Connectivity ...... a video capture card or dongle can offer a range of video inputs (and outputs in some cases). The simplest capture devices use a composite video signal. This is the lowest quality signal that you will find on AV equipment. That said, it is likely what you will have to work with on thermal cameras that provide an analogue video output ! For information, in terms of quality this is the list from worst to best.....

Composite Video - Y/C (S-Video) - RGB - Component - HDMI

SDI will be in the area of HDMI I suspect and is found on professional camera equipment.
Maybe add Cameralink on top. Huge bandwidth for raw thermal data but the capture cards and software are $$$ :(

6. Monochrome video - This week I used an EZCap USB 2.0 dongle to capture video from a monochrome output SWIR camera. The dongle is actually a decent model but it did not like the monochrome video input and claimed no signal was present when I tried to record. I fooled the units input checks using a colour bar chart video input from my CCTV test monitor before swapping the signal to the camera output. It was a faf to do and highlights that modern video capture chipsets MAY have an issue with missing parts of a signal that they expect to find, such as the colour burst. Video capture devices offer several Television Signal standards in their input configuration but on my dongle all were colour and there was no option for a monochrome PAL or NTSC signal. I know that other capture dongles do not have this issue so it may be uncommon or limited to a certain chipset. Some may not check the video input for anything except the sync pulses.
One thing I did not (yet) test. Will be included in the PAL test.

7. Video standards conversion...... thermal imaging cameras commonly present either a PAL or NTSC signal at their video output socket. The video capture device needs to be set for the video standard that is present. All of my capture devices support PAL, NTSC and SECAM without issues. The video standard can be changed in software when the recording is transferred to a DVD, Memory Stick etc. There is not normally a need to change the video standard with a hardware solution between the camera and the video capture device.

8. Video link standard conversion..... this is a new one for me so I can offer no in depth experience beyond a single test I carried out recently. I had a DVD recorder that provided only component, SCART RGB and SCART Composite video output. I needed an HDMI signal to feed into our new television. I bought a “Techole” SCART to HDMI converter from Amazon. It had good reviews. I installed the converter and was impressed that it was truly ‘plug and play’ with only the output resolution of 720P or 1080P to be selected with a push button. The unit up scaled the input video signal :) Now the bad news. Whilst the converter worked well, I did some research and discovered that most inexpensive SCART to HDMI converters use a chip set that only accepts a composite video input, even if they have a SCART socket. This means that a video source that can provide RGB with better quality cannot deliver that improved image to the televisions HDMI input. More expensive converters will take component or RGB video inputs. Again, with a thermal camera the issue is moot as the output signal is often plain old composite video. In my case, I ‘dumped’ the elderly DVD recorder and bought a wonderful used Sony RDR-HXD870 HDD/DVD recorder that provides an upscaled 1080P HDMI output  ;D Great recordings and great image quality on our new TV.
The SCART to HDMI converter I used was originally purchased to connect an old Amiga 600 to a new TV. To get best image quality I bought one which was stated to work with separate RGB signals (and it did  ;D). Note: To change the input from Composite/CVBS to RGB, pin 16 of the SCART connector needs to supply ~1-3V.

Using a camcorder as a capture device ...... For many years the UK camcorder market offered Camcorders that could only output line video and not record it. I believe this was due to the different Duty rates between line out only camcorders and video recorders as a camcorder with a video line input was classed as a VCR. You could buy the same model of camcorder overseas that had the video line input capability enabled ! I bought such camcorders but, in truth, rarely used the line input recording feature. My last camcorder was a Canon DV tape based unit and that had the line input recording enabled. The Camcorder could record analogue video from a video source and play that digitally recorded video back via its FireWire port to a host PC for editing etc. It could also do live real time conversion from analogue video source to DV output from its FireWire port. Is this a good option in place of a modern USB video capture device ? Well it is a solution but there are issues. The quality of video conversion may, or may not be as good as that of a dedicated USB video capture device. The resolution of video capture may also be relatively low. The video standard of the input to the camcorder may be fixed so not suitable for other video standard recordings. The Camcorder may have an inactivity shutdown feature that switches it off if not actually recording. FireWire is sadly now obsolete so less common on modern PC’s and laptops. It can still be added via a PCIe card or Expresscard however. The camcorder is a more bulky and complicated solution than buying a decent dedicated video capture device.
An other good idea, maybe I can get a DV or other camcorder as portable capture device.
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
EzCap "EasierCAP" USB capture device
« Reply #4 on: January 09, 2021, 04:47:51 am »
EzCap "EasierCAP" USB capture device
STK1160 based with GM7113C as ADC. No AC97 Audio codec (only 8kHz Sound of the STK).
Bought at Aliexpress for 6,43$ https://www.aliexpress.com/item/32629043559.html

Is recognized as "USB 2.0 Video Capture Controller" without driver and as "STK1160 ATV BDA" and "USB Audio Interface" with diriver. I used Windows 7 x64 and the drivers from http://www.logilink.de/download/driver/VG0001.zip. The drivers are maybe not the best but work  :D
VID: 05E1 PID: 0408

The device supports these resolutions:
YUY2: 160x120, 176x144, 320x240, 352x288, 640x480, 720x480
I420: 160x120, 176x144, 320x240, 352x288, 640x480

Display/recording software was VLC (as always).

Outside


PCB top


PCB bot
« Last Edit: January 09, 2021, 04:53:57 am by Cat »
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
EzCap "EasierCAP" USB capture device
« Reply #5 on: January 09, 2021, 04:49:05 am »
I tested these settings:
Planar 4:2:0 YUV (I420)
640x480 30FPS
Bitrate: 110592kb/s
Image is cropped.



https://drive.google.com/file/d/1l0Jf8Wye972rhFZbdMNNs2In6xb1WShP/view?usp=sharing

Packed YUV 4:2:2, Y:U:Y:V (YUY2)
640x480 30FPS
Bitrate: 147456kb/s
Image is cropped.



https://drive.google.com/file/d/1r5pVq42tiV-NjpuktmgOd4-o3dKFNmvt/view?usp=sharing

Packed YUV 4:2:2, Y:U:Y:V (YUY2)
720x480 30FPS
Bitrate: 165888kb/s
Image is uncropped but with a blue line on the right.



https://drive.google.com/file/d/1-K8EAjACKGpKHxnaEfGsVN6MJcGHuVUI/view?usp=sharing
« Last Edit: January 10, 2021, 06:15:31 pm by Cat »
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
UVC USB2.0 Video capture device
« Reply #6 on: January 09, 2021, 05:10:08 am »
UVC USB2.0 Video capture device
ARK3399 (V399) based with AMT630A as ADC.
Bought at Aliexpress for 6,26$ https://www.aliexpress.com/item/32716198483.html

Is recognized as "USB2.0 PC CAMERA" and "USB2.0 MIC".
No driver is needed thanks to UVC https://en.wikipedia.org/wiki/USB_video_device_class
VID: 18EC PID: 5555

The device supports one resolution:
MJPEG: 640x480


Display/recording software was VLC (as always).

Outside


PCB top


PCB bot



I tested these settings:
Motion JPEG Video (MJPG) Planar 4:2:2 YUV full scale
640x480 30FPS
Bitrate: 8314kb/s
Image is slightly cropped on the top and bottom, the left side has a dark line. Image appears a bit darker.



https://files.catbox.moe/klp9qt.avi
https://drive.google.com/file/d/1KNv4fUMiWyXocOt7TpOf8CSvVmUA6yZV/view?usp=sharing
« Last Edit: January 10, 2021, 06:09:21 pm by Cat »
On the Internet, nobody knows you're a cat.
 

Offline _Wim_

  • Super Contributor
  • ***
  • Posts: 1524
  • Country: be
Re: UVC USB2.0 Video capture device
« Reply #7 on: January 09, 2021, 08:10:01 am »
Image appears a bit darker.

If you put a series attenuator or a Y-adaptor with 50-ohm terminator on the easier cap devices, the image is probably also darker and in my case,  the contrast was quite a bit better also.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Re: Analog video capture device comparison
« Reply #8 on: January 09, 2021, 11:14:20 am »
If you put a series attenuator or a Y-adaptor with 50-ohm terminator on the easier cap devices, the image is probably also darker and in my case,  the contrast was quite a bit better also.
Thank you, I'll give it a try.
DC input impedance was around 75 Ohm for all devices but maybe I'll check the complex impedance as well.
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Medion MD85921 USB video grabber
« Reply #9 on: January 09, 2021, 11:34:31 am »
Medion MD85921 USB video grabber
TM5600 based (Copyright on the chip 2004 ::))
Device was sold at Aldi supermarkets in 2008 and later.

Is recognized as "TridVid Capture" and "Trident Analog Video" with driver from http://www.medion.com/de/service/_lightbox/handbuch_details.php?did=6947
VID: 6000 PID: 0001

The device supports these resolutions:
UYVY: 352x288, 352x576, 720x576 (reported for PAL with 25 FPS)
When recording NTSC the image captured was 720x480 with 30FPS.

Display/recording software was VLC (as always).

Outside


Label


PCB top


I tested these settings:
Packed YUV 4:2:2, U:Y:V:Y (UYVY)
720x480 30FPS
Bitrate: 165722kb/s



https://drive.google.com/file/d/1j72iidejxqWDpZn6xpIfoiZplYOsLklI/view?usp=sharing

When there is no signal present blue and green lines flicker on the screen

« Last Edit: January 10, 2021, 06:09:51 pm by Cat »
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
SCART to HDMI converter + HDMI to USB UVC converter
« Reply #10 on: January 09, 2021, 12:12:42 pm »
SCART to HDMI converter + HDMI to USB UVC converter
The main IC of the SCART to HDMI converter has a removed marking  >:(
The HDMI Transmitter is an IT66121.
Bought some years ago for ~20$, unfortunately I can't find where. It was stated to support RGB input as well (for old gaming consoles).
The Device can be switched between 720p and 1080p HDMI output.
With NTSC input nothing else can be changed.
With PAL RGB input also 50Hz and 60Hz could be selected.

To capture the HDMI there are inexpensive USB2 capture cards
Bought at Aliexpress for 7,24$ https://www.aliexpress.com/item/4001051182130.html

Is recognized as "USB Video" and "USB Digital Audio".
No driver is needed thanks to UVC https://en.wikipedia.org/wiki/USB_video_device_class
VID: 534D PID: 2109

The device supports many different resolutions:
MJPG:  640x480, 720x480, 720x576, 800x600, 1024x768, 1280x720, 1280x960, 1280x1024, 1360x768, 1600x1200, 1920x1080 all with 30FPS
YUY2: 640x480, 720x480, 720x576, 800x600, 1024x768, 1280x720, 1280x960, 1280x1024, 1360x768, 1600x1200, 1920x1080 (different FPS for each resolution, 30FPS for 640 or 10FPS for 1280)

SCART to HDMI converter
Outside






PCB



HDMI to USB UVC converter
Outside



I tested these settings:
HDMI capture card: Motion JPEG Video (MJPG) Planar 4:2:2 YUV full scale
1280x720 60FPS
Bitrate: 63646kb/s
SCART to HDMI Converter set to 720p.



https://drive.google.com/file/d/1jL1l7pFZboQPAqBUbMybNOJcFth1mYCg/view?usp=sharing

The overlay takes some time to disappear. The image quality is impressive (fine detail at the ® symbol, the whole range of the video level is used), unfortunately there are a few drawbacks. The image is heavily cropped (at least with composite NTSC) and the aspect ratio is 16:9 (because of the selected resolution of the HDMI capture card).

With no input signal a blue screen with the overlay is displayed.

« Last Edit: January 10, 2021, 06:10:22 pm by Cat »
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Canopus ADVC-110 IEEE 1394 capture device
« Reply #11 on: January 09, 2021, 12:40:37 pm »
Canopus ADVC-110 IEEE 1394 capture device
Semiprofessional capture and playback device with SAA7115 as ADC.
Got it some time ago from a HAM.

Of course you need a IEEE 1394 interface, I used an Expresscard with two IEEE 1394 ports for my laptop.
No driver was needed for this setup. When connected the Canopus is recognized as "AV/C Tape Recorder/Player" with following IDs: VEN 2011 MOD 0 TYP 4

Input signal type is set by a DIP switch, nothing can be changed by software.

Display/recording software was VLC (as always).

Outside








PCB


DIP-switch settings

On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Re: Canopus ADVC-110 IEEE 1394 capture device
« Reply #12 on: January 09, 2021, 12:50:15 pm »
I tested these settings:
DV Video (dv  ) Planar 4:1:1 YUV
720x480 30FPS
Buffer size: 736x480
Bitrate: 30307kb/s
Setup level set to 0 IRE



https://drive.google.com/file/d/1VSko_u1fvfO0jpM1F1cIwYBanv1Kz_fS/view?usp=sharing

No cropping but black borders on the right and left side. Transitions with high contrast are flickering slightly.

DV Video (dv  ) Planar 4:1:1 YUV
720x480 30FPS
Buffer size: 736x480
Bitrate: 30307kb/s
Setup level set to 7,5 IRE




https://drive.google.com/file/d/103kJS8U72BgSMkjQqPIFe2-bfHAYt8gV/view?usp=sharing

Like above but with a different black level.
« Last Edit: January 10, 2021, 06:11:26 pm by Cat »
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Eachine ProDVR
« Reply #13 on: January 09, 2021, 01:10:07 pm »
Eachine ProDVR
The first standalone solution is this Eachine ProDVR, normally used for FPV quadcopter stuff.
TVP5150AM1 ADC.

Bought at Banggood for 12,60€ (CZ warehouse) https://www.banggood.com/Eachine-ProDVR-Pro-DVR-Video-Audio-Mini-Recorder-for-FPV-Multicopters-for-RC-Drone-FPV-Racing-p-1061196.html
The device records and plays from a microSD-card (FAT32, max. 32GB), three different recording resoultions can be set with the menu. A monitor connected to the video output is needed for this. The video output can be changed from NTSC to PAL, the recording status is overlayed on this output. Once everything is set up only one button is needed to start the recording.

On the PCB are two SMPS, some other forum posts suggest that the noise in the image can be reduced when replacing them with LDOs. When Idle the input current is 153mA @ 5V and during "HD" recording the current increases to 163mA.

Outside


Cables are included as well (without RCA connectors)


PCB


« Last Edit: January 10, 2021, 06:11:50 pm by Cat »
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Re: Eachine ProDVR
« Reply #14 on: January 09, 2021, 01:25:37 pm »
Recording quality set to "HD"
Motion JPEG Video (MJPG) Planar 4:2:0 YUV full scale
1280x720 30FPS
Bitrate varies between 11 and 17 Mb/s





Slight cropping, wrong aspect ratio.
Directly from SD: https://files.catbox.moe/h5uoow.AVI
https://drive.google.com/file/d/1FcOjbqQ-O3dKTt89ujrKp12djrmBqebh/view?usp=sharing

Recording quality set to "D1"
Motion JPEG Video (MJPG) Planar 4:2:0 YUV full scale
720x480 30FPS
Bitrate varies between 2 and 7 Mb/s




Slight cropping.
Directly from SD: https://files.catbox.moe/58bu7e.AVI
https://drive.google.com/file/d/1KFsJpWIQopLDavH3gVqisvh63EjCA7qv/view?usp=sharing


Recording quality set to "VGA"
Motion JPEG Video (MJPG) Planar 4:2:0 YUV full scale
640x480 30FPS
Bitrate varies between 4 and 6,6 Mb/s





Slight cropping.
Directly from SD: https://files.catbox.moe/tsweik.AVI
https://drive.google.com/file/d/1PGcNUrTt6oqjw3C05T3LayZ3rMCFEodw/view?usp=sharing

I'll change the firmware of the device after I could test it with PAL. There are forum posts indicating the newer firmware fixed PAL cropping issues.
« Last Edit: January 10, 2021, 06:13:30 pm by Cat »
On the Internet, nobody knows you're a cat.
 

Offline Vipitis

  • Frequent Contributor
  • **
  • Posts: 867
  • Country: de
  • aspiring thermal photography enthusiast
Re: Analog video capture device comparison
« Reply #15 on: January 09, 2021, 01:49:47 pm »
I will have to read through most of this later. May I suggest a tool called YUView. It lets you analyse video compression. While you can see macro blocking and patterns with some filters. There is a way to load multiple sequences and do a difference mask. For me the effective bitrate (read in the bitstream analysis) is the most important as it's directly proportional to the quality of the image. Also the keyframe interval matters on interframe encoding.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
HMDVR-S
« Reply #16 on: January 09, 2021, 01:53:38 pm »
HMDVR-S
The second standalone solution is this HMDVR-S, normally used for FPV quadcopter stuff.
Like the ProDVR with TVP5150AM1 ADC.

Bought at Aliexpress for 10,92$  https://www.aliexpress.com/item/32915558924.html
The device is very small and lightweight, records to microSD-card (FAT32, max. 32GB).
No setup is possible, only start and stop recording with a button.

On the PCB are many LDOs and no SMPS. One LDO got very hot (~95°C) so I added the orange silicone pad to spread the heat. When Idle the input current is 225mA @ 5V and during recording the current increases to 239mA. Forum posts indicate the device can be operated with ~3,5V and thus stays much cooler.

Outside


PCB




Recorded with the only available setting.
Motion JPEG Video (MJPG) Planar 4:2:0 YUV full scale
1280x720 30FPS
Bitrate varies between 13 and 20 Mb/s




Slight cropping, wrong aspect ratio.

File directly from the SD-card: https://files.catbox.moe/0ey97f.AVI
https://drive.google.com/file/d/12NQ5Hxr8koLb6KrQsqY1PwQ0fmOReU6Z/view?usp=sharing
« Last Edit: January 10, 2021, 06:13:54 pm by Cat »
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Re: Analog video capture device comparison
« Reply #17 on: January 09, 2021, 02:05:54 pm »
May I suggest a tool called YUView. It lets you analyse video compression. While you can see macro blocking and patterns with some filters. There is a way to load multiple sequences and do a difference mask. For me the effective bitrate (read in the bitstream analysis) is the most important as it's directly proportional to the quality of the image. Also the keyframe interval matters on interframe encoding.
Thank you, I'll have a look at YUView. Additionally I'll upload my test clips.
For the PC based solutions I used VLC to record as (AVI container). From the two standalone devices the files are directly from the SD-card.

I'm not quite sure if bitrate is everything, it seems like the ADCs or the settings of them limit the performance of the standalone solutions. It seems like the analog bandwidth is limited and thus the ® symbol not really recognizable.

UVC devices can be easily recorded with an android mobile phone, so maybe this (SCART to HDMI ->HDMI to USB) is the way I'll go...
« Last Edit: January 09, 2021, 02:09:14 pm by Cat »
On the Internet, nobody knows you're a cat.
 

Offline CatTopic starter

  • Regular Contributor
  • *
  • Posts: 96
  • Country: de
Re: Analog video capture device comparison
« Reply #18 on: January 09, 2021, 02:27:09 pm »
A quick note, this was my test sequence
  • Let the Scout run and start recording
  • Restard the Scout
  • Wave fingers infront of the Scout
Of course the fingers are out of focus but give a "feeling" about the framerate.
On the Internet, nobody knows you're a cat.
 

Offline Fraser

  • Super Contributor
  • ***
  • Posts: 13170
  • Country: gb
Re: Analog video capture device comparison
« Reply #19 on: January 09, 2021, 02:34:04 pm »
Cat,

Thank you for the time and effort that you are investing in these tests. I know from experience that such testing can take a lot of time to complete.

 :-+ :-+

Fraser
If I have helped you please consider a donation : https://gofund.me/c86b0a2c
 

Offline Vipitis

  • Frequent Contributor
  • **
  • Posts: 867
  • Country: de
  • aspiring thermal photography enthusiast
Re: Analog video capture device comparison
« Reply #20 on: January 09, 2021, 05:21:31 pm »
While we are on the topic, I want to quickly detail what a professional solution would consist off. If you get lucky and find and older external recorder that takes AV in, like an Video Devices PIX24 you can use that to record high bitrate ProRes of a analog source. Nowadays you would need to buy a converter like one from Blackmagic (easily 200bucks), power it with 12V and add a recorder that takes the HDMI or SDI signal out. For a whole working solution I got quoted 1300€ which included the converter, cables, a recorder (Video Assist also from BMD) and a portable power solution using a V lock plate.
If you don't need an external monitor you could just go with the BMD h246 pro recorder for 470€ which still needs to get powered.

If you are at home there are a few other options for capture cards that take the full signal.

But you have to understand that any camera will do their own Digital to analog conversion as the initial signal from the sensor isn't PAL or NTSC resolution nor interlaced - which changes the whole discussion...
 

Offline _Wim_

  • Super Contributor
  • ***
  • Posts: 1524
  • Country: be
Re: SCART to HDMI converter + HDMI to USB UVC converter
« Reply #21 on: January 09, 2021, 05:44:16 pm »
SCART to HDMI converter + HDMI to USB UVC converter
The main IC of the SCART to HDMI converter has a removed marking  >:(
The HDMI Transmitter is an IT66121.
Bought some years ago for ~20$, unfortunately I can't find where. It was stated to support RGB input as well (for old gaming consoles).

You can still buy it: https://nl.aliexpress.com/item/32988881670.html?gps-id=6311502&scm=1007.24625.129096.0&scm_id=1007.24625.129096.0&scm-url=1007.24625.129096.0&pvid=79444c03-a8d8-47a0-add3-ec0919410dea&spm=a2g0o.store_home.fullPieceDiscountPromo_6000067974706.32988881670

 
The following users thanked this post: Fraser

Offline _Wim_

  • Super Contributor
  • ***
  • Posts: 1524
  • Country: be
Re: Analog video capture device comparison
« Reply #22 on: January 09, 2021, 07:58:41 pm »
Just discovered my DFG/USB2PRO is an older DFG/USB2-It based on a Tv master tm5600 chipset. They look identical, and as the label of my device was covered by a company sticker, I made the wrong conclusion is was a DFG/USB2PRO (as this was the only device mentioned on the imaging source website). Sorry for the confusion.
 

Offline Bill W

  • Super Contributor
  • ***
  • Posts: 1103
  • Country: gb
    • Fire TICS
Re: Analog video capture device comparison
« Reply #23 on: January 09, 2021, 08:31:21 pm »
OK time to add my 3.

EzCap (China copy with 'Forward Video' branding)
VID=EB1A PID=2861 driver/source recognition as 'USB2861'
USB device view full of errors about 'invalid InterfaceSubClass'

Hauppauge Win-TV HVR1100 card (BT88x) PCI card

WinTV USB-Live2
VID=2040 PID=C200 driver/source recognition as 'CX23100'

Software (record)
Videoview (Syntek Directshow 2004)
VLC (2.2.8)
Win-TV8 (for WinTV products only)

Software (Playback)
MPlayer
VLC
Win-TV8

First up I had a bit of a problem with software co-existence as both 'videoview' and VLC often would not pick up the WinTV devices as they seemed to get 'attached' to the native WinTV8 software if it had run before on that power cycle of the PC.  Equally WinTV8 would then ignore the devices / channels.

Add to that VLC changing colours when doing a screenshot and deciding the camera is PAL if asked to play sources ... may have to come back to this  |O

For now then just some stills grabbed with MPlayer.
Camera is an Argus 'Frankencamera' in NTSC, with a ULIS 320x240 ASi (sort of).

Of note the WinTV-USB has a blue line on the left and has a much harder time dealing with the interlace - see the 'fur' on green battery bar outline which is one pixel/TV line wide.
The other two from are 'Videoview' recordings 15fps RGB uncompressed as YUV 4:2:2 avi 480 TVL


Bill
 
The following users thanked this post: _Wim_

Offline _Wim_

  • Super Contributor
  • ***
  • Posts: 1524
  • Country: be
Re: Analog video capture device comparison
« Reply #24 on: January 10, 2021, 07:47:21 am »
Of note the WinTV-USB has a blue line on the left and has a much harder time dealing with the interlace - see the 'fur' on green battery bar outline which is one pixel/TV line wide.

That surprised me quite a bit, as this uses the same chipset as the semi-professional DFG/USB2PRO for which I had high hopes:
https://www.linuxtv.org/wiki/index.php/Conexant_CX2310x

 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf