Lossless HDMI capture devices - comparison screenshots

[Brad]

Extreme pixel-peep edition!

Only comparing 3 of my 5 devices, at least for now:

• StarTech USB3HDCAP | Driver: "1.1.0.160.6, 2015.11.23"
• Blackmagic Design Intensity Pro 4K | "Blackmagic Desktop Video" version: v10.5.4
• AVerMedia C027 (AVerTV HD DVR) | Driver: 1.12.0.22

480i YCbCr 4:2:2




The "480i" IP4K captures were cropped by 3px on top and 3px on bottom: HDMI only supports 480 active lines for 480i/p but the IP4K always captures 486i for each. The extraneous lines are just pure Y=16 black.

The C027 and IP4K captures of the 480i 4:2:2 Menu + Info screen are bit-identical (confirmed by SHA-1 hash of direct PNG screengrabs).

I have no explanation for the USB3HDCAP's deviations. Anyone wanna take a stab?

The only thing I can think of is that it offers Proc Amp controls that actually alter the signal when they're modified (even Sharpness). But these "should" leave the picture unaltered when left at default. In comparison, while the C027 has Proc Amp controls for Brightness/Contrast/Hue/Saturation, all are non-functional for HDMI sources. The IP4K only offers adjustments for analog sources.

1080p/i YCbCr "4:4:4" (source content is actually 4:2:0)

(This image is just to confirm that the PS3's YCbCr output is 4:4:4 not 4:2:2 -- unfortunate for capture purposes.)



None of the capture devices support capturing to YCbCr 4:4:4, so they all have to downscale the chroma horizontally to 50%. I'm really surprised the differences in scaling are this visible when the source was only 4:2:0 to begin with.

The IP4K's chroma is horizontally aliased. (But only horizontally, right? They should all be the same vertically, in theory, and that seems to pan out to my eyes.)

I don't see any luma deviations with the USB3HDCAP here, unlike the 480i 4:2:2 captures.

The PS3 places its info overlay in a slightly different location for the 1080i output, and deliberately blurs it. Not that this matters.

Update 2016-07-03:

480p YCbCr 4:4:4



Greyscale() -> PNG -> SHA-1 hash confirms that the luma of these C027 and IP4K 480p captures is identical. (The 480i note above regarding cropping and IP4K's reinterlacing of 480p to 486i applies here, mind you.)

As with the 1080 4:4:4 captures, we see the different chroma downscaling implementations at play. Yet, the IP4K's "V" channel is blurry and horizontally-shifted in the second comparison?

[Cauptain]

Hello vaporeon800,

You could make a full review about Blackmagic Design Intensity Pro 4K capturing PS3 at 1080p@60fps?

I'm interested in bought and would like to know if compensated

Can u record a 1 minute video to see the quality?




Claudio

[Brad]

1080p60 HDMI capture works now, but even the current drivers have annoying bugs. Like sometimes needing to unplug & replug my HDMI cable from the splitter when a signal format change occurs.

Even a few seconds of lossless 1080p60 is too unwieldy to upload. An RGB capture looks as good as it would look on your TV screen, so what is it you're hoping to see with a sample?

[Cauptain]

I want see a 1080p@60fps to compare with my 1080p@30 C985 records converting to 60fps with spirton interframe script.

Whats games u have?

If u can, record and encode using x264 at 12k. Give 40~50Mb file.





Claudio

[Cauptain]

double post!!

[Brad]

Cauptain, did you receive my PM (Jan 19)?

[Cauptain]

Cauptain, did you receive my PM (Jan 19)?

I received and answered, however as it was at work dont know if the message was sent.

Anyway, I have this game and if you can record quality that I said, will be good because I bought a new 4K TV and I want to see how it comes out also in upscaler.

Thanks mate



Claudio

[Brad]

Sorry for the super long delay. I was going to try capturing a new sample, but after some issues I just decided to encode a lossless file I had sitting around.

Can u record a 1 minute video to see the quality?

If u can, record and encode using x264 at 12k. Give 40~50Mb file.

It's 1 minute at ~12000 kbps, but that doesn't work out to the filesize you expected. Anyway, for this type of content, the bitrate is too low to avoid obvious compression artifacts -- at least when conforming to Level 4.2.



Blackmagic IP4K 1080p59.94 sample - Super Stardust HD demo screen.avs

AVISource("2015-07-21 @ 12-xx - Super Stardust HD demo screen [ULRG].avi") # RGB32 Full range (0-255)
AssumeFPS(60000,1001) # Capture header from VDub is 10000000/166833 = 59.94017970... rather than 59.94005994... Difference is 1 frame every ~8350 secs = >2 hrs.

ConvertToYV12(matrix="rec709") # Compress to 16-235 (no dither)

Encode Video (1080p60 level 4.2 CRF placebo no tune).bat

set filepath=Blackmagic IP4K 1080p59.94 sample - Super Stardust HD demo screen
set crf=29.5

x264-r2665-a01e339 --colorprim "bt709" --transfer "bt709" --colormatrix "bt709" --preset placebo --sar 1:1 --level 4.2 --crf %crf% "%filepath%.avs" -o "%filepath% V.mkv"

MediaInfo

Duration                       : 59s 693ms
Overall bit rate               : 12.2 Mbps

Video
ID                             : 1
Format                         : AVC
Format/Info                    : Advanced Video Codec
Format profile                 : High@L4.2
Format settings, CABAC         : Yes
Format settings, ReFrames      : 4 frames
Codec ID                       : V_MPEG4/ISO/AVC
Duration                       : 59s 693ms
Bit rate                       : 11.9 Mbps
Width                          : 1 920 pixels
Height                         : 1 080 pixels
Display aspect ratio           : 16:9
Frame rate mode                : Constant
Frame rate                     : 59.940 fps
Color space                    : YUV
Chroma subsampling             : 4:2:0
Bit depth                      : 8 bits
Scan type                      : Progressive
Bits/(Pixel*Frame)             : 0.096
Stream size                    : 84.8 MiB (98%)
Writing library                : x264 core 148 r2665 a01e339
Encoding settings              : cabac=1 / ref=4 / deblock=1:0:0 / analyse=0x3:0x133 / me=tesa / subme=11 / psy=1 / psy_rd=1.00:0.00 / mixed_ref=1 / me_range=24 / chroma_me=1 / trellis=2 / 8x8dct=1 / cqm=0 / deadzone=21,11 / fast_pskip=0 / chroma_qp_offset=-2 / threads=6 / lookahead_threads=1 / sliced_threads=0 / nr=0 / decimate=1 / interlaced=0 / bluray_compat=0 / constrained_intra=0 / bframes=16 / b_pyramid=2 / b_adapt=2 / b_bias=0 / direct=3 / weightb=1 / open_gop=0 / weightp=2 / keyint=250 / keyint_min=25 / scenecut=40 / intra_refresh=0 / rc_lookahead=60 / rc=crf / mbtree=1 / crf=29.5 / qcomp=0.60 / qpmin=0 / qpmax=69 / qpstep=4 / ip_ratio=1.40 / aq=1:1.00
Language                       : English
Default                        : Yes
Forced                         : No
Color range                    : Limited
Color primaries                : BT.709
Transfer characteristics       : BT.709
Matrix coefficients            : BT.709

[Brad]

Updated the OP with a little 480p 4:4:4 comparison.

[dvgeek]

Hello vaporeon800 - I'm planning to buy the Blackmagic Design Intensity Pro 4K - but have lots of questions - would it be possible to give some answers please?

- What software do you use for capture?
- Does the Media Express software do capture?
- Using Codec's or Lossless?
- If using codec's - are they selectable or fixed?
- What about audio codecs? If I have a directshow DTS encoder codec, can I use that? I'm assuming the audio comes in as LPCM. If it is not LPCM, can it capture the bitstream?

Thanks.

[Brad]

- What software do you use for capture?

I prefer VirtualDub and lossless capture.

- Does the Media Express software do capture?
- Using Codec's or Lossless?
- If using codec's - are they selectable or fixed?

Media Express can capture. Last time I used it, the options were only uncompressed or MJPEG, with no access to external codecs. Their page currently says that it can also do DNxHD, DVCPRO, "and more".

- What about audio codecs? If I have a directshow DTS encoder codec, can I use that? I'm assuming the audio comes in as LPCM. If it is not LPCM, can it capture the bitstream?

Received and captured as LPCM. Others have reported success with getting different Blackmagic cards to capture the bitstreams packed as "noise" into the LPCM data and then decoding it in software, but I wasn't able to do this with the IP4K.