A consumer-grade LCD monitor for precise visual stimulation

1
Dear Mario and all.

I remember that someone in the forum was asking questions about appropriate LCD monitors that could replace the CRTs, so you may be interested in this.

Recently psychologists are looking for a suitable new mode display (e.g., LCD) that can replace the CRT display. Although there are two suitable monitors on the market (Display++ and VPixx ), both monitors are very expensive.

One of our recent studies has found a consumer-grade monitor (Asus PG278Q，or PG279) that can be used for precise visual stimulation. When we set it in a specific mode (ULMB mode) and after proper software correction (via Psychotoolbox), the performance is comparable to the CRT monitor in property of temporal and spatial homogeneity. Moreover, different from the CRT's progressive scan, the target monitor (under the ULMB mode) presents the whole screen stimuli at the same time. This may be of great implication for improving the signal-noise ratio (SNR) of ERPs data.

The full text can be download either via https://link.springer.com/article/10.3758%2Fs13428-018-1018-7 or https://raw.githubusercontent.com/yangzhangpsy/papers/master/beh2018.pdf.

Best wishes
-YANG
2
Thanks Yang.

I assume you use the default pulse width of "100" when taking the ULMB measurements?

Some notes to any researchers who may happen upon this post:

ULMB stands for ultra-low motion blur which is an NVIDIA-branded moniker for strobing the backlight once per refresh. The general idea here is that strobing the backlight on sample-and-hold LCDs mimics the rasterizing found on CRTs. You can read more about reducing motion blur here [1,2].

Unfortunately, ULMB is unavailable on a non-NVIDIA graphics source. If you're an AMD or Intel user, then your options are limited to native blur-reduction monitors from Asus, BenQ, or Eizo (and maybe others by now).

As a general note to others who need to replace CRTs in the near future: familiarize yourself with trade offs between various types of LCD panel technologies (TN, IPS, and VA are the big three) and find a thorough review site to help you decide on the model. The good sites should measure spatial homogeneity, pixel rise and fall times, and motion blur reduction. I personally like TFTCentral.

[1] http://www.tftcentral.co.uk/articles/motion_blur.htm
[2] https://www.blurbusters.com/faq/motion-blur-reduction/
3

Very interesting paper and discussion of strobing backlight technique.

 

I have a small question regarding the “temporal properties” section of your results. When transitioning from a black frame to a white frame (RML1st data), you observed a 19% (98.6-79.5) difference in luminance between the top of the display and the bottom. Presumably this is because the lower pixels have had less time to transition than the higher pixels. You propose below that this unequal luminance be corrected via Psychtoolbox. How would this work? It seems to me that the luminance correction is not just a function of pixel location, but also a function of the data from previous video frames. For example, if the previous frame was black, then the bottom pixel might need a 19% increase in video value; but if the same pixel were white in the previous frame, then it would require less correction, or maybe none at all if the pixel has been white for multiple frames. How would we correct for this pixel memory effect in software?

 

-Peter

 

 

From: PSYCHTOOLBOX@yahoogroups.com <PSYCHTOOLBOX@yahoogroups.com>
Sent: Monday, March 26, 2018 9:22 PM
To: PSYCHTOOLBOX@yahoogroups.com
Subject: Re: [PSYCHTOOLBOX] Re: A consumer-grade LCD monitor for precise visual stimulation

 

 

Dear Zack，

Thanks for bringing us the detail of the ULMB technique. 

 

Yes, currently it is a little bit inconvenience that the ULMB is only available in nVidia-branded monitor + Nvidia cards. However, the price for those devices is acceptable for most of the lab in this field (around 300 and 700 USD for the GTX 1060 3G graphics card and the ASUS PG279Q monitor respectively). As a comparison, both CRS Display++ and Vpixx are much much more expensive than this (nVidia card plus the ASUS monitor for a total price of around 1000 USD).

 

Besides, as can be seen in our paper, we have thorough tests on the spatial homogeneity, temporal homogeneity, pixel rise, hold, and fall times. After software correction via Psychtoolbox, all of those characters are comparable or even superior to the CRT monitor(fig 7). 

 

Moreover, different from the CRT's progressive scan, the target monitor (under the ULMB mode) presents the whole screen stimuli at the same time. This character makes it a preferred monitor for ERPs labs who wanna improve further the signal-noise ratio (SNR) of ERPs data.

 

Best

-Yang

 

On Tue, Mar 27, 2018 at 7:23 AM, zack_lb@... [PSYCHTOOLBOX] <PSYCHTOOLBOX@yahoogroups.com> wrote:

 

Thanks Yang.

I assume you use the default pulse width of "100" when taking the ULMB measurements?

Some notes to any researchers who may happen upon this post:

ULMB stands for ultra-low motion blur which is an NVIDIA-branded moniker for strobing the backlight once per refresh. The general idea here is that strobing the backlight on sample-and-hold LCDs mimics the rasterizing found on CRTs. You can read more about reducing motion blur here [1,2].

Unfortunately, ULMB is unavailable on a non-NVIDIA graphics source. If you're an AMD or Intel user, then your options are limited to native blur-reduction monitors from Asus, BenQ, or Eizo (and maybe others by now).

As a general note to others who need to replace CRTs in the near future: familiarize yourself with trade offs between various types of LCD panel technologies (TN, IPS, and VA are the big three) and find a thorough review site to help you decide on the model. The good sites should measure spatial homogeneity, pixel rise and fall times, and motion blur reduction. I personally like TFTCentral.

[1] http://www.tftcentral.co.uk/articles/motion_blur.htm
[2] https://www.blurbusters.com/faq/motion-blur-reduction/



 

--

 

Yang Zhang

------------------------------------------------

Department of Psychology

Soochow University

Dushu Lake Campus

No.1 Wenjin Street, Industrial park

Suzhou, Jiangsu, 215123, China

 

     Or: zhangyang873@...

     Or: yangzhangpsy@163.com

------------------------------------------------