So, I'm digging about for info on something unrelated to my 2408 problems, and I run across this on Wikipedia:
Dell 2407WFP... had major banding issues and other display quality control issues before the A04 revision in which Dell disabled a video processing chip inside of the 2407WFP when in "desktop" mode.
Oh man, really? This is true? There was a "desktop mode" that switched off the processing inside the 2407?
If so, this would suggest that Dell learned nothing from their experiences with the previous 24" model. :smileysad:
Please, pretty please, add a "desktop mode" to the 2408 so we can banish the sharpening (and probably the lag) issues on the 2408. And let me swap my 2408 when you've done it!
Right, but "Desktop" on the 2408 is simply a colour balance preset, isn't it? Doesn't seem to do anything other than switch to some rather dubious colour rendering.
Tell me you don't have to put up with a preset colour balance on the 2407 just to switch the processing off?
I imagine that the 'Desktop' setting is the neutral default, with no adjustments. The alternatives are 'Multimedia' or 'Gaming'. It seems to me that the
ACC/ACM processing was not meant to be enabled in 'Desktop' mode in the first place.
If all the Genesis does is colour re-mapping, I don't see how it would cause any significant lag or sharpness problems.
I think it's whatever does the sharpening and scaling that causes the lag (and the "sharpening" issues). For instance, in the movie world, scaler-deinterlacer chips are well known for causing a frame or two's delay to video, because of the time it takes to process the data. That would correspond to about 20-40ms of lag. That may not sound like much, but it's not difficult to see when video is one frame out of sync with audio, and slipping by two frames is horribly obvious.
I haven't had to deal with this stuff for a while, but from what I remember, I think Genesis make scaler chips too.
the Genesis link talks about about color space conversions.
"In order to enhance the color and contrast ACM transforms the video signal from the native red/green/blue (RGB) color pace into the luminance/chrominance (YUV) color space"
I noticed an apparent improvement to lag using the HDMI input vs. the DVI input, i.e. I never measured it. After retesting, it does not seem like a dramatic improvement. I don't use HDMI regularly because I can't get the colors quite right.
But it does make me wonder if the 2408 has this technology somewhere in the processing path. And if so, is this processing done twice, i.e. back to RGB once theG enesis finishes color correction.
Clearly the 2408 has to create RGB data to drive the panel, since LCD panels are RGB devices. Whether that's done by the same chip that converts to YUV and does the mapping I don't know - don't even know for sure if it uses that Genesis chip. But I don't think colour conversion and mapping are likely to cause much lag, because it's just swapping one pixel value for another. It's scaling and sharpening that are processor intensive (and tend to produce artefacts like the ones we're seeing).
When no scaling or sharpening are required i.e. when feeding 1920 x 1200 to the monitor digitally, it should be possible to bypass the processing. If we could, I'm certain most of the issues with the 2408 would go away.
time for me bench some of the monitor settings. will use my 2005FPW on the DVI, maybe HDMI too. may help to isolate the lag, who knows. i'll to to post anything of interest on this thread tonight.
day off today, so I did some benching of 2408WFP versus 2005FPW, full details further below.
my observations:
- different preset modes and sharpness showed no real differences in lag
- defeated sharpness could not be tested, unknown method
- scaling not thoroughly tested (i do not scale), focus was presets
- avg all 2408 DVI testing was 51.25 msec behind the 2005
- avg all 2408 HDMI testing was 19.96 msec behind the 2005
- HDMI via adapter on my configuration was faster, on avg by 31.26 msec
- what the author at monitortest.blogspot.com alluded to, appears to be correct
raw data below, no real use posting pictures somewhere. btw, the (tiny) 2005FPW is back in the box
=========
- both monitors 1680 by 1050, 2005 is filled, 2408 is centered with black bars, ATI h/w scaling off
- numbers are least significant millisecs, format is {2408WFP, 2005FPW, absolute diff} - captured with canon xsi ISO 1600 1/2000th sec, all photos dark but very clear - ati cat 8.451, all defaults, clone mode, crossfire disabled - 2 asus 3870 crossfired using 2 bridges - 2408 on DVI with "graphics" setting, custom RGB at {92, 90, 98} with brightness 20 contrast 50
- wide mode 1:1 mappingunles specified
- 2005 on DVI, brightness 100, contrast n/a, sharpness n/a - sharpness 50 unless otherwise listed. Varied, not defeated (unknown method) - used test from http://www.lagom.nl/lcd-test/ with FF3, no load on machine, Vista64, Q6600@ 3.4, 4GB DDR2 @ 1133 - any inconclusive pictures dropped
asus 3870 with supplied dvi-to-hdmi adapter, on the card. then hdmi cable between adapter and 2408.
btw, this hdmi adapter (according to asus) allows digital audio to go over the HDMI, if you connect a digital audio source onto the card (i.e. within the case). i haven't done it, and it should have no impact, but maybe worth mentioning here.
gpro
2 Intern
•
2024 Posts
791
0
Posted August 1st, 2008 01:00